The wind turbine blade is the core component of the wind turbine. However, wind turbine blades will suffer fatigue and internal crack defects, which seriously affect the safety service performance in harsh environments. The current detection mode is mainly traditional manual detection and shutdown maintenance, with high risk and low efficiency. Therefore, this research proposed a method to conduct a non-stop online inspection of the wind turbine blade’s internal defects using ultrasonic inspection mounted unmanned aerial vehicle. As the vital component of this inspection system, the optimized design of the negative pressure adsorption car was carried out. The structural characteristics of the negative pressure adsorption car were determined experimentally, followed by multivariate analysis and an integrated development and performance test study of the test system. Based on the test results, the detection rate of this test system for internal wind power blade faults is 100%. This research offers novel concepts of and theoretical insights into the intelligent and security identification of internal flaws in wind turbine blades.